disorderedmaterials/dissolve

Significant figures on charges

Closed this issue · 1 comments

Describe the bug
It is possible to import, for example, a LigParGen forcefield containing charges and get a legitimately zero-charged configuration (within the specified tolerance). However, on saving that input file the charges get truncated and on re-loading the configuration is no longer zero within tolerance.

Charges are saved with 6 d.p. precision, so are arguably "accurate enough". How to enforce the neutrality is tricky, and the failure to run is probably more informative and less dangerous, so closing as a non-issue.